A Novel Source Code Clone Detection Method Based on Dual-GCN and IVHFS

Source code clone detection, which can identify fragments with similar functions, plays a significant role in software development and quality assurance. Existing methods either extract single syntactic or semantic information, ignore the associated information between statements different structures. It is difficult for these to effectively detect pairs functions. In this paper, we propose new model based on dual graph convolutional network (GCN) interval-valued hesitant fuzzy set (IVHFS), named DG-IVHFS. Specifically, simplified grouped abstract syntax tree (AST) of source obtain group representations. The representations AST, as well control flow (CFG) representations, were transformed into structures, then applied GCNs them learn dependencies nodes. addition, introduced IVHFS more comprehensive evaluation similarity. Our experimental results demonstrated that precision, recall, F1-scores DG-IVHFS BigCloneBench GoogleCodeJam datasets reached 98, 97 97% 93 95%, respectively, exceeding current state-of-the-art models. Moreover, our performed terms time consumption.